# Implementation suggestion my use case

**URL:** <https://discuss.huggingface.co/t/implementation-suggestion-my-use-case/77265>\
**Category:** 🤗Transformers\
**Created:** [March 13, 2024, 1:30pm UTC](https://discuss.huggingface.co/t/implementation-suggestion-my-use-case/77265 "2024-03-13T13:30:12Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Sam-2002](https://avatars.discourse-cdn.com/v4/letter/s/4491bb/32.png) [@Sam-2002](https://discuss.huggingface.co/u/Sam-2002)\
**Post date:** [March 13, 2024, 1:30pm UTC](https://discuss.huggingface.co/t/implementation-suggestion-my-use-case/77265/1 "2024-03-13T13:30:12Z")

</div>

Hello,

I am working on a recommendation system and a part of it takes a user’s bio and recommends entities based on the entities’ descriptions. As you can understand, this isn’t just text matching or semantic search as the two(bio and description) might differ a lot on the basis of terms but still have a significant correlation based on the context.

I was looking into Elastic’s ELSER model but realized that isn’t a good fit. How can I implement this and which transformer and model would be great for my use case? I am new to HuggingFace so I would appreciate any help!

Thanks a lot!